Đang chuẩn bị liên kết để tải về tài liệu:
Key frame extraction from videos - a survey

This paper carries out an extensive survey of different key frame extraction techniques along with their merits and demerits used in current video retrieval application. | ISSN:2249-5789 Azra Nasreen et al , International Journal of Computer Science & Communication Networks,Vol 3(3),194-198 Key Frame Extraction from Videos - A Survey Azra Nasreen Assistant Professor Dept of Computer Science & Engineering RVCE, Bangalore. Azranasreen@rvce.edu.in Dr. Shobha G Professor & Dean Dept of Computer Science & Engineering RVCE, Bangalore. shobhag@rvce.edu.in Abstract With the advent of social networking, sharing of multimedia has gained tremendous amount of importance and is the widely used form of communication worldwide. In the process of discovering knowledge from videos, challenge is to process huge amount of information which is resourceintensive. One way to minimize the cost of computations is to reduce the amount of information that undergoes processing. The redundant information can be greatly reduced by extracting key frames that represent the entire video. It is an essential task in any video analysis and indexing applications. This paper carries out an extensive survey of different key frame extraction techniques along with their merits and demerits used in current video retrieval applications. 1. Introduction Video contains huge amount of information at different levels in terms of scenes, shots and frames. To discover knowledge from videos the issue that needs to be addressed is the elimination of redundant information. The Objective is to remove the redundant data which will significantly reduce the amount of information that needs to be processed. So, key frame extraction is the fundamental step in any of the video retrieval applications. It is necessary to discard the frames with repetitive or redundant information during the extraction. In recent years, many algorithms of key frame extraction focused on original video stream have been proposed. This paper provides an extensive survey in this area to bring out the advantages, drawbacks, suitability to an application, and precision of each method for video retrieval .